Comparative analysis of the penetrative capacity of synopses and of full papers unrelated to the synopses published in the same broad-scope agricultural journal.

As from 1984, synopses based on unpublished research reports have been published in the Netherlands Journal of Agricultural Science (NJAS), in addition to full papers unre lated to the synopses.To study readers' behaviour towards the synopsis, which is still a relatively uncommon vehicle of primary communication, both full papers (n = 94) and synopses (n = 67) published in NJAS in 1984-1986 were subjected to citation analysis. Self-ci tations ("autocitations," as opposed to "allocitations") were excluded from most analyses. On average, citation yields were significantly lower for synopses than for full papers. The citation yields vaned strongly between the three years.The distribution of the journal's contents over the various subdisciplines of agricultural science explamed fairly well the capricious behaviour of citation yields over the years, but not the difference in yield between synopses and full papers.Further, the coverage of NJAS papers by five major biblio graphic databases was analysed. AGRIS and SCISEARCH covered NJAS' contents integrally. Thirteen percent of all papers (n = 21) covered by AGRIS and SCISEARCH only had not been cited at all. CHEMABS' coverage appeared to be fmrly consistent with the scope of this database (chemistry). However, BIOSIS' and CABI's coverage behaviour over the period in question appeared to be characterized by incon sistency and arbitrariness. Coverage by BIOSIS and CABI was significantly better for full papers than for synopses (BIOSIS 87% and 43%. and CABI 65% and 48%, respectively).After correction for bibliographic coverage the citation yield was still lower for synopses than for full papers, but the difference was no longer significant. Two-thirds of the initial difference in citation yield between full papers and synopses could be attributed to differences in bibliographic coverage.It is concluded that inconsistent bibliographic coverage procedures seriously undermine the penetrative capacity (and hence meaningfulness) of papers published in journals whose readers depend largely on bibliographic sources. [ABSTRACT FROM PUBLISHER]